side note: I think alot of hashing power is staying on Core because it seem controversial to "undermine the core devs". this most incorrect way of thinking needs to stop. competing implementation are perfectly valid and should be welcome.
1. This is not the web in which you can choose Safari, Mozilla or Chrom(e/ium) where these 3 programs all operate under the same web / http protocol.
2. The "competing implementations" in our case, is "
your web sucks, I'll have my own web - along with my own browser".
3. Satoshi didn't even welcome ...compatible implementations, let alone competition of incompatible ones.
PS. The web evolves by something like a "...soft-forking" of new web clients and web servers who support new features like the HTML5 video player... Those running older browsers can access content without these features but not content with it (they'll have to revert to a flash player in this example). They can always upgrade to get access to newer features but the web is always the same. There are no two webs.
It is consensus, not the software you run decide what is bitcoin. If core changed total number of coins to 42 million, would you still use core software because you want one web?
At what criteria do you define a new feature to be rule breaking? Segwit soft fork demonstrated by using soft forks you can always do rule-breaking changes (like change the transaction/block format) while still not being rejected by the old clients. That change is rule breaking thus it is a hard fork, just sold as a soft fork